Friday’s BLS truther controversy was, in a word, sad. That folks now nonchalantly float claims that government agencies fudge numbers is (or should be) beyond the pale (just as it was in 1970 when Nixon did it). But it’s not. While the truther discussion has made its way into every nook and cranny of the interwebs, maybe it would be instructive to take a look at some anecdotal information that supports Friday’s allegedly contrived unemployment number.

Business Insider’s Joe Weisenthal suggested early Friday that the numbers we’d gotten earlier last week on auto sales foretold a decent jobs/unemployment number, which sent me scurrying to FRED to produce a chart:

(Note: Unemployment Rate is inverted to better show correlation. Chart is current through Friday’s release)


The pattern of auto sales leading the unemployment rate is crystal clear. Great call, Joe. So, the same folks who are skeptical of the BLS must necessarily also question the sales reports of GM, Ford, Toyota, Honda, BMW, and every other car manufacturer that reports monthly sales figures (i.e. all of them).

Another piece of anecdotal information that I’ve used for years now is the NFIB’s “Poor Sales” (as Single Biggest Problem) vs. the Unemployment Rate:


As of the NFIB’s most recent report, Poor Sales was cited by 20 percent of businesses as their Single Biggest Problem in their Small Business Economic Trends report.

Finally, the BLS itself goes to great pains to be transparent, and does a good job doing so. Earlier this year, their Editor’s Desk column ran a piece titled Employment Trends From Two Surveys, which very broadly discussed the two surveys from which jobs data is compiled.

Separately, and almost certainly unknown to Mr. Welch, the BLS issues a technical document every month that address the trends in both surveys. That document can be found here, and contains the following chart:



Among the objectives of the monthly analysis is to produce an Adjusted Household Survey (seen above). BLS:

This [Adjusted Household Survey] is a research series created from household survey employment to be more similar in concept and definition to payroll survey employment. Household survey employment is adjusted by subtracting agriculture and related employment, nonagricultural self employed, unpaid family workers, private household workers, and workers absent without pay from their jobs, and then adding nonagricultural wage and salary multiple jobholders. The effects of population control revisions also have been smoothed out in the historical data in this series.

But you knew that, right, Jack? The Adjusted Household Survey shows that 1.836MM jobs have been created over the past 12 months. The Payroll Survey? 1.806MM. The difference between the two over the course of a year? A paltry 30,000 (see Page 1 of the PDF, numbers below). While there may occasionally be wide month-to-month swings (and the Household Survey is known to be the more volatile of the two), the two series generally track fairly closely over time, which is how data should generally be observed.

And where were the conspiracy theorists earlier this year when the Adjusted Household Survey was flagging 400k+ job losses? Nowhere, that’s where.


Source: BLS, Adjusted Household Survey Level (Column 3), MoM Adjusted Household Survey Change (Column 4), MoM Nonfarm Payroll Survey Change (Column 5)

The notion that career BLS economists, statisticians, and staffers were collectively in the tank to produce a result designed to influence the election is absurd on its face. It’s a waste of everyone’s time to discuss and debate this. I’ve seen no one make a claim that this has been a robust recovery; it has not been, and it continues to be tepid. That said, we should be channeling our efforts on how to make it better and not making idiotic and irresponsible claims that there’s widespread data book-cooking going on.

Lastly, I’ll confess, too, that I’m having a bit of difficulty confirming Mr. Welch’s oft-repeated claim that we added 600,000 government sector jobs in the past two months. Can anyone help me out here, or was that also pulled from somewhere within his bowels? I’m always open to the possibility that I’ve overlooked something, and if that’s the case, I’d love to see what it was – I can find no evidence that the government added 600k jobs in the past two months.

Adding: Thanks to Jeff in Indy for the link to Mish, who points to Table A-8 (Household Data) of the release for data on the 600k bump in government employees:

Now, what’s above is Seasonally Adjusted, so seasonality is out of the picture here. That said, there are clearly outliers in this series fairly often (Jan 2003 jumps right out at me, as does April 2012′s -442. And there’s no readily apparent pattern that I can discern). Again, I would stress that data of this type needs to be observed over time – it’s typically not instructive to draw inferences from one or two months’ prints.

Here are the Establishment and Household surveys on Government Employees over time.  As a fellow (fairly prominent) econ blogger just emailed me: “Nobody would use the household survey to track the number of government employees.” Indeed, and I don’t know anyone who has (which is why I was at a loss as to what Welch was talking about). That said, I’ll be inquiring as to the gap that opened – and persisted – since the early 90s.

Category: Current Affairs, Data Analysis, Economy, Employment, Really, really bad calls

Please use the comments to demonstrate your own ignorance, unfamiliarity with empirical data and lack of respect for scientific knowledge. Be sure to create straw men and argue against things I have neither said nor implied. If you could repeat previously discredited memes or steer the conversation into irrelevant, off topic discussions, it would be appreciated. Lastly, kindly forgo all civility in your discourse . . . you are, after all, anonymous.

35 Responses to “Addressing the BLS Truther Controversy”

  1. I like to remind people that George W. Bush tried to illegally fire 7 US Attorneys for political reasons. It burst into view and became a scandal — a mere 7 Attorneys!

    Anyone thinks that the White House could doctor the BLS report without it leaking out simply is not rooted in any form of reality.

  2. Argyle says:

    Why are we even paying attention to a doddering old man obviously off his meds?

    Dementia is a sad thing — please don’t make fun of him, just wheel him back to his room and move on.

  3. HP5976 says:

    Invictus, is clearly a card carrying Dem. And that is fine. However to state that unemployment has dropped by .3 points in a month is pure foolishness. As the operator of a temporary staffing firm in 8 states it is NOT true.
    Check tax receipts, it is easy to find the truth and the BLS under R or D is BS. [BR: You mean these? They were up 5.3% in Q3]

    And to suggest that we are not subject to abuse by Government statistics is also foolish, just check out ANY predictions on the budget, the FED etc.

    Conspiracy?? No criminality with John Corzine? The Government clearly picks winners/losers in todays fully rigged HFT market.


    Invictus: So, just to be clear here – when, at the end of the Bush administration/early Obama months, the Unemployment Rate rose by 0.4, 0.3, 0.5, 0.5, 0.5, 0.4, that was bullshit, too. Right? Are you on board for that? When it went from 6.8 to 9.4 in a matter of months (mostly under Obama), that was bullshit, right? Cause it just can’t move in such “large” increments, right?

  4. jeff in indy says:

    here’s mish’s take. also mentions the 680k increase in govt jobs:


  5. Orange14 says:

    @Invictus – Most school teachers work on 10 month contracts so they are out of the labor market for July and August and back in September (my daughter is one and she’s still trying to learn how to budget with 10 paychecks rather than 12). Maybe this helps explain that one.

  6. RW says:

    I have heard that former-CEO’s, particularly those who profited greatly from manipulating their company’s books, are not only unindicted criminals but sociopaths and pathological liars who periodically engage in bestiality with sundry barnyard animals. I find this quite plausible but, no, I have no evidence; I’m just raising the question.

  7. Frilton Miedman says:

    Another variable that supports the BLS numbers, IRS payroll revenues. – this site – http://www.dailyjobsupdate.com/ predicted a decent jobs report earlier in the week based on it.

    I’m sick of hearing so many on the right make so many constant & ridiculous assertions, there comes a time when empirical data trumps political deception & hyperbole, now is that time.

    Jack Welch has removed any semblance of personal credibility this week, the extremists in the GOP are only reinforcing their growing reputation for ridiculous hyperbole & lies, the same public opinion that has Congress at a 10% approval rating.

    To Jack Welch and the other idiots who support his claims, get over it, this game is over and you’ve lost…accept it and move on or get left behind by the rest of America.

  8. Greg0658 says:

    wondering on those auto sales / or auto leases and the flow rates .. does the government (& farm sector) go shopping with the fresh budget (harvest) balance?

  9. Hey Barry,

    The 604 government jobs come from table A-9, and so is a very volatile series. We thought about mentioning that, but didn’t want to promote that series because of the noise factor. In that series private non-ag employment is up just 50K over the last two months. And Orange14 makes a good point–there could be some double counting of teachers.

  10. DrungoHazewood says:

    Its weird with an approval rating of 10 and a reelection rate of over 80. Who’s fault is that? Why the people doing the disapproving. And enjoy the senile rantings of the criminal Welch as he attacks his fellow criminals. And let’s fire up the FIRE economy once again as it eats the nation for lunch as people scream for more debt. Its all such a good show.

  11. brianinla says:

    What are we looking at – two years of positive job growth? If the trend is so positive why did the latest report show those on food stamps at record highs? Or are the jobs created so worthless the people also qualify for a government handout? Another year from now will we be at <7% unemployment but 48 million on food stamps? And most important if we are at those numbers would you still consider it a positive trend?

  12. JDJ says:

    So the BLS lied? Is that because if the latest crop of “truthers” worked for the BLS, its what they would do?

    “There is a cult of ignorance in the United States, and there has always been. The strain of anti-intellectualism has been a constant thread winding its way through our political and cultural life, nurtured by the false notion that democracy means that “my ignorance is just as good as your knowledge.”
    ― Isaac Asimov

  13. econimonium says:

    You know, if these people would suddenly wake up (OK, they won’t but if the people actually running the Republican Party did) they’d realize that this is the exact reason they’re losing the election and why, with the economy the way it is, Romney isn’t doing better. It’s because these crazy people have driven all the sane and moderate Republicans, like me, out of the party and now we vote Democratic. I will not, ever, vote for a Republican until the party is cleared of the right-wing crazies, conspiracy theorists and religious nuts and makes a turn toward reality. And here in MA, people like me would have given Brown another term. Now? I’d just be supporting the crazies in Washington if he got re-elected. And let me tell you, it’s the primary reason Warren is now tracking better. The more these people talk like this, the worse the party’s chances get. Unless, of course, you live somewhere that you’ve been taught creationism is real. Then you deserve the stupid you get.

  14. Conan says:

    Interesting that the U3 number was reduced by .3 to 7.8, however a more true and not talked about number is U6. This number remains unchanged at 14.7.

    An even better number to use, in my opinion cuts through a lot of BS, is Labor Force Participation Rate and it is at 63.9% or about where we were in 1981 and just an up tick of .1% from our latest low.

    So I’m not going to go the conspiracy route, but what I am going to say is we’re talking about the wrong number. U6 is better to use than U3 and Labor Force Participation is actually the bottom line for all of it.

  15. A7LB says:

    Interesting that U-6 unemployment remained at 14.7%…

  16. Molesworth says:

    Ditto econimonium. Lifelong Rep but have found myself voting Dem bec of Rep angry crazies. Haven’t switched or gone Indep bec I like to get both sets of propaganda as my spouse is Dem.

  17. There are few people who have been advocating the reportage of U6 longer than I; see: Unemployment Reporting: A Modest Proposal (U3 + U6)

    But understand this: U3 (and 4, 5) and U6 all report different things.

    U6 is a measure of UNDER-Employment versus U3 UN-Employment. They measure different things, move in parallel but not lockstep.

  18. contrabandista13 says:

    Everything is a bloody conspiracy nowadays… Jack Welsh, I never liked that guy, he’s a cheap stuffed suit who fancies himself to be some kind of autocratic oracle.. Every time I see him on CNBC, the first thought that comes to mind is “short dick, white trash”…. It’s just so typical for him to be a “Truther”, I’m not the least bit surprised, I wouldn’t have expected anything else from him…..

  19. RW says:

    BTW, Catherine Rampell does a good job of explaining the volatility in the jobs number series and, eo ipso, the most likely reason(s) for the jump in numbers. (ht Dean Baker)

    Conspiracy theories are usually a function of ignorance, of incomplete knowledge, and they often ramp up in volume when the stakes are raised. Team Republican does not view favorable economic data as favorable to its purposes, that is all, so when a good number pops up then that needs to be poo-pooed but if it pops up in an improving trend (no matter how anemic) then that number needs to be chopped and its validity trashed.

  20. Lee Adler says:

    Invictus does an excellent job here. Perhaps the best evidence is the 6.3% gain in withholding taxes y/y in Sept, which works out to 2.3% real, even stronger than the bump in the household survey. http://wp.me/p2r1d8-v8r

  21. jnkowens says:

    I’m with you econimonium. I am a lifelong republican who cannot believe the insanity that passes for reason in the GOP these days. Rove and friends opened a pandora’s box when they whored themselves to the tea partiers and now they’ve probably lost a winnable election. I will cast my vote for Obama this November, not because he’s earned it, but because the other guy represents an outcome that i find horrifying.

    If you’ve ever sat in a history class or watched a documentary and wondered how a nation could let a party so full of venemous hatred as the Nazi’s rise to power, witness the conspiracy theories, xenophobia, homophobia, racism, and misogyny emanating from the right wing of the republican party and wonder no more.

    It’s so sad that it’s come to this.

  22. Northeaster says:

    The only BLS model that matters:


    While in steady decline, the time series didn’t start going up until women started entering the workforce, then in earnest in roughly 1980.

    So what? If this BLS model is accurate to filter out conspiracy theories, then there simply are not enough jobs to keep up with population growth. People with no jobs, or low income jobs, don’t pay into a system that has a growing population more reliant on that system. Not hard science here.

  23. carleric says:

    If the Republicans can ever be considered a viable political party then they are going to have to come up with workable solutions to our problems….something the liberals also fail on although they seem to appeal to the worse elements of human nature. Look, no serious person gives a damn about the issues that are at the forefront of this laughable political campaign. Personally I became a moderate anarchist years ago and am thinking of dropping the moderate tag. Vote? For what and why? Like George Carlin I won’t be leaving my house on election day.

  24. ilsm says:

    Neutron Jack is a bomber, he turned GE into a financial asset.

    Jack never worried that he sold engines for military airplanes which are for fighting WW II again.

    But he likley relates BLS truthiness to the miilitary contracts he profited from shoddy deliveries, and no one ever lost money selling to Uncle Sam.

  25. Frilton Miedman says:

    jnkowens Says:
    October 7th, 2012 at 7:39 pm
    “I’m with you econimonium. I am a lifelong republican who cannot believe the insanity that passes for reason in the GOP these days. Rove and friends opened a pandora’s box when they whored themselves ……”

    “If you’ve ever sat in a history class or watched a documentary and wondered how a nation could let a party so full of venemous hatred as the Nazi’s rise to power, witness the conspiracy theories, xenophobia, homophobia, racism, and misogyny emanating from the right wing of the republican party and wonder no more.

    It’s so sad that it’s come to this.”


    As a Libertarian, I COMPLETELY agree, I’m sickened to see the mantra of “personal freedom” of the Libertarian ideology somehow become perverted into “corporations are people”.

    Political bribery labelled “campaign contribution” combined with a monumental brainwashing/propaganda campaign paid for by the likes of Karl Rove and the Koch brothers, similar to the way Hitler gained power behind the scenes in the 1930′s – promoting public fear and using it to attain greater and greater power over democracy until there is no democracy left.

  26. MikeG says:

    Welch is projecting onto the BLS what he would do in the same situation – lie, cheat and manipulate the numbers to what he wants. This arrogant, sociopathic camera hog needs to STFU, no-one except the Rabies Right is buying your PR garbage anymore.

  27. ski3938 says:

    If Welch was still the head of GE he never would have commented on anything
    publicly 5 seconds after he heard it. He would of had it studied by his
    ad visors and lawyers and maybe issue some kind of statement after it
    was vetted by his staff. But now he is a senile egotistic rich loudmouth
    and he not only shoots from the hip but actually admits he has zero
    proof.. If any one of his staff ever did go public with this nonsense
    they would have been fired fast. Welch worked all his life to make money
    and a reputation and was willing to lose it in 5 seconds!!!

  28. victor says:

    Had the number gone up to say 8.4%, the Romney Campaign would be celebrating and the Obama Campaign would be doing a lot of damage control, hand wringing and bloviating. Jack Welch would be radiant under that scenario, praising BLS for such a good job. Truth is the employment scene is not good in general and piss poor in California and Nevada where I do business. By the way if you want to find Gov. manipulation of employment numbers look a bit farther away, look to China. But, a trend is a trend albeit an anemic one and meantime the ones taking it in the chin are of course the needy ones.

  29. SecondLook says:

    A couple of data points:


    Labor force Men Women

    1950 86.4 33.9
    1960 83.3 37.7
    1970 79.7 43.3
    1980 77.4 51.5
    1990 76.4 57.5
    2000 74.8 59.9
    2005 73.3 59.3
    2010 72.6 59.7

    Total percentage of the labor force participating:

    1950: 59.2
    1960: 59.4
    1970: 60.4
    1980: 63.8
    1990: 66.4
    1998: 67.1 (peak year)
    2012: 63.6

    Two things are obvious, one that women entering in large numbers the workforce starting in the 1970′s contributed largely to growth of the participation rate.
    Second, that the rate per se, isn’t indicative of the health and prosperity of the country. Consider the rates between 1950-1970; a period that most would call the most prosperous in terms of employment and income growth in modern history.

    What really matters is how wages are doing, and there lies the problem. They aren’t doing well at all. Since 1975 real wages have been essentially stagnant. Real total median household income rose (which includes all incomes, and other revenue streams such as interest, dividends), by less than 20% net, over the past 4 decades.

    As for U-6, as noted above, it’s not a measure of unemployment, but more of the quality of employment. One of the weakness of the U-3 report is that it counts part time workers, even those who only work a one (!) day a month. The fact that U-6 didn’t change while U-3 slightly improved suggests that taking on part-timers was a factor in the latter.
    However, part-time is better than no time, and that means, however modest, households are getting some more cash flow. Which we all should agree, is a good thing.

  30. Jim67545 says:

    If one is a facile and unconcerned liar, it is natural that one would think that the other guy is lying too. Romney’s and Ryan’s refusal to recant some of the outright, baldfaced lies in their own ads in my mind flatly disqualified them. If someone is quite willing to lie, to say anything expedient to get elected (as in the debate), he will lie and say anything expedient when/if in office. I don’t understand why this fundamental moral issue (not to mention religious issue where lying is still one of those 10 commandments – that “false witness” one) is not more foremost in peoples’ minds. Lying is not to be dismissed as crafty politics. It’s a fundamental moral flaw.

  31. Fake Jobs Numbers Would Look Better Than This

    When the Bureau of Labor Statistics (BLS) reported on Friday that the unemployment rate had fallen to 7.8% in September, some observers wondered if the numbers had been “cooked” for political purposes. They can relax. Fake jobs numbers wouldn’t look as bad as these.

    Two numbers in the BLS “Employment Situation” report attracted concern: the reported 873,000 increase in total employment, and the 0.3 percentage point reduction in the unemployment rate. These figures suggest a rapidly improving labor market, which would be very convenient for President Obama right now. However, as soon as one delves deeper into the BLS numbers, the reality of continued economic stagnation becomes clear.

    As the White House has said repeatedly (and correctly), it isn’t good to read too much into any one month’s employment numbers. So, let’s look at the third quarter of 2012 as a whole.

  32. 10x25mm says:

    Ultimately, the problem here is that the two different surveys in the NFP report have created jobs numbers that are way outside of each other’s statistical confidence intervals. This would not be surprising in the typical political poll with a sample size around 1,000, but it should never happen in polls with 40,000 to 50,000 size samples. Either the Establishment Survey or the Household Survey – or both – suffer from a significant error. The BLS should have directly addressed this massive discrepancy in their NFP report, not just stated that the jobs number were derived from different polls.

    This is why the NFP report was met with considerable disbelief and it is not fair to label those who are calling out this discrepancy ‘truthers’. Something is clearly wrong here and the BLS is not explaining what happened.

  33. philipat says:

    There you go again!!

    I would suggest that, to avoid such future distractions, we simply refer to U-6 as a better, more consistent , measure of “under-employment” compared to The Household survey and also more compatible with the U-3 data for the purposes of comparison. The recent divergence between the trend lines of U-3 and U-6 then becomes more signivicant as the basis for discussion. But I am not holding my breath for Invictus to engage in a discussion of this devergence any time soon.